Implementation of GraphNe Network and fuzzyPN optimization in human resource allocation

This study combines Fuzzy optimization and Graph Neural Network (GNN) to improve
decision-making in Human Resource Management (HRM). When it comes to assigning
employees to tasks, evaluating performance, and making recruitment selection, here we
use Fuzzy pentagonal numbers to offer a more easy way to express linguistic evaluations
like “poor, average, good, very good, and excellent” criteria’s. We use the HRM system as
a heterogeneous graph, where employees, tasks, and departments are represented as nodes,
while the relationships between skills and workload dependencies are the edges. Each node
and edge is enriched with FPN-based attributes to handle uncertainty in evaluations and
preferences. So we propose a hybrid framework that combines Graph Neural Networks
(GNNs) with fuzzy pentagonal number (FPN)-based optimization. Finally, an FPNbased
multi-objective optimization model is applied to derive optimal HRM decisions
using Sigmoid Function such as employee-task assignments, maximizing skill-task fit and
minimizing workload imbalance. This integrated approach enables more realistic, datadriven,
and uncertainty-aware decision-making in HRM systems. Numerical example will
illustrate the study and represent the optimization based on the frame work taken.